BpmrufmHnwagi

We

Ub

ASm

N

0d

Safety and Effectiveness of Granulocyte-Colony Stimulating Factor in MobilizingStem Cells and Improving Cytokine Profile in Advanced Chronic Heart Failure

Jacob Joseph, MDa,*, Asem Rimawi, MDb, Paulette Mehta, MDb, Michele Cottler-Fox, MDc,Anjan Sinha, MDb, Balkrishnaman K. Singh, MDb, Rebecca Pacheco, BSNb,

Eugene S. Smith III, MDb, and Jawahar L. Mehta, MD, PhDb,d

The objective of this study was to determine a safe and effective dose of granulocyte-colony stimulating factor (G-CSF) to mobilize hematopoietic stem cells in patientswith advanced heart failure and to determine its effects on the cytokine profile.Patients with advanced heart failure (n � 6) and implantable defibrillators in situwere administered G-CSF after baseline echocardiographic and laboratory evalua-tion, using an escalating dose schedule designed to ensure safety. The peripheralCD34� hematopoietic stem cell count increased from 3.6 � 0.5/�l to 38.7 � 13/�l(p � 0.022) after 5 days of 5 �g/kg/day G-CSF therapy. The baseline or peak whiteblood cell count did not predict the stem cell response. G-CSF increased plasma levelsof interleukin-10. Left ventricular ejection fraction increased significantly in the 4patients with ischemic cardiomyopathy 9 months after treatment. No major adverseeffects attributable to the drug occurred during administration or 9 months offollow-up. Our results have shown that low-dose G-CSF significantly mobilizedhematopoietic stem cells in advanced heart failure and improved left ventricularfunction in the ischemic subset of patients. G-CSF significantly increased plasmalevels of the anti-inflammatory cytokine interleukin-10, without changing pro-inflammatory cytokine levels. In conclusion, these results indicate a novel mech-anism of action for the potential therapeutic benefit of G-CSF in advancedischemic cardiomyopathy. © 2006 Elsevier Inc. All rights reserved. (Am J

Cardiol 2006;97:681– 684)

ffdtamAupClGfldltimtibwre

one marrow stem cell transplantation and mobilization areromising therapies for chronic heart failure (HF). Initial ani-al studies in infarct models have demonstrated myocardial

egeneration with hematopoietic stem cell (HSC) mobilizationsing growth factors.1 Recent clinical studies have shown theeasibility and potential clinical benefit of autologous bonearrow cell transplantation in HF.2–4 Stem cell mobilization inF is an attractive alternative to the more cumbersome tech-ique of stem cell transplantation. However, it is not knownhether stem cells can be safely mobilized in patients with

dvanced HF, especially because high doses of multiplerowth factors are required for stem cell mobilization in clin-cal practice and in preclinical studies.1,5,6

• • •e enrolled patients followed at the Central Arkansas Vet-

rans Healthcare System who had left ventricular ejection

aDepartment of Medicine, VA Boston Healthcare System and Bostonniversity School of Medicine, Boston, Massachusetts; Departments of

Medicine, cPathology, and dPhysiology and Biophysics, University ofrkansas for Medical Sciences and Central Arkansas Veterans Healthcareystem, Little Rock, Arkansas. Manuscript received June 15, 2005; revisedanuscript received and accepted September 19, 2005.

Dr. Joseph was supported by Grant 2M01RR014288-060041 from theational Institutes of Health, Bethesda, Maryland.

* Corresponding author: Tel: 857-203-6841; fax: 857-203-5549.

bE-mail address: jacob.joseph@med.va.gov (J. Joseph).

002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.oi:10.1016/j.amjcard.2005.09.112

raction �35%, advanced (New York Heart Associationunctional class III to IV) HF, implantable cardioverterefibrillators in situ (to avoid arrhythmic complications dueo HSC engraftment and potential electrical heterogeneity),nd had been maintained on stable HF therapy for �3onths. The institutional review board of the University ofrkansas for Medical Sciences approved all proceduressed in this study. Baseline laboratory measurements wereerformed before study drug administration, includingD34� cell number by flow cytometry, plasma cytokine

evels, and echocardiography. Patients were admitted to theeneral Clinical Research Center of University of Arkansas

or Medical Sciences and given granulocyte-colony stimu-ating factor (G-CSF) at an initial dose of 2.5 mg/kg twiceaily subcutaneously for 5 days, with careful clinical andaboratory monitoring. The study was designed to increasehe dose in the absence of demonstrable efficacy with thenitial dose (Figure 1). Serial laboratory testing and clinical

onitoring were conducted daily during drug administra-ion, at days 6, 8, and 10, at 6 weeks, and at 3 months afternitiation of treatment. Echocardiograms were recorded ataseline, 5 and 10 days after the start of therapy, and at 6eeks, 3 months, and 9 months. The echocardiograms were

ead by 2 cardiologists blinded to the study. The primarynd point was an increase in CD34� cells in the peripheral

lood higher than the predetermined efficacy level of 10

www.AJConline.org

Ccccl(CSCflv(ctb

Tdm3qetitdbtCt�wNnw(

rtdmf

cetsionb

piwwtiwaof

Mctbmcciepcptmptw

cstnGcititirtce

Flcc

682 The American Journal of Cardiology (www.AJConline.org)

D34� cells/�l (a level that ensures adequate collection ofells for transplantation). The secondary end points werehanges in left ventricular ejection fraction, safety, andhanges in plasma cytokine levels. Plasma levels of inter-eukins (ILs)-2, -4, -5, and -10, tumor necrosis factor-�TNF-�) and interferon-� were assessed by the Cytokineore Facility at the University of Arkansas for Medicalciences using BD Cytometric Bead Array Human Th1/Th2ytokine kit (BD Biosciences, San Diego, California) andow cytometry. The paired t test was used to comparearious parameters before and after therapy using SigmastatSPSS, Inc., Chicago, Illinois), with a p value of �0.05onsidered significant. Pearson’s product moment correla-ion coefficient was computed to analyze the correlationetween variables.

The study was conducted in groups of 3 each (Figure 1).he initial group of patients who were given the startingose of 2.5 �g/kg twice daily responded with adequateobilization of HSCs; hence, we repeated the same dose inmore patients. Because all 6 patients responded ade-

uately, the trial was terminated without additional dosescalation. The patients were older, had severe left ven-ricular systolic dysfunction, and included an octogenar-an with advanced HF (Table 1). Five patients received aotal dose of 22.5 (n � 2) to 25 �g/kg (n � 3) within 5ays, and 1 patient received a total dose of 10 �g/kgefore the study drug was stopped because of an eleva-ion in alkaline phosphatase (Table 1). The peripheralD34� cell count increased in all patients to higher than

he predetermined efficacy parameter of 10 CD34� cells/l. The mean CD34� cell count increased significantly,ith a parallel increase in the white blood cell count.either the baseline white blood cell count nor the mag-itude of increase in the white blood cell count correlatedith the CD34� cell response in an individual patient

Table 2).The intent-to-treat analysis, including the patient who

eceived only a total dose of 10 �g/kg G-CSF, showed thathe left ventricular ejection fraction measured by echocar-iography was increased by an absolute value of 6.3% at 9onths (Table 3). The increase in left ventricular ejection

igure 1. Study design. An escalating dose design was used to determineowest dose for safe and effective mobilization of hematopoietic stemells. *Effective: leading to peripheral CD34� cell count of �10ells/�l. BID � twice daily.

raction was significant in the 4 patients with ischemic n

ardiomyopathy. The left ventricular end-systolic andnd-diastolic dimensions tended to decrease after G-CSFreatment, but these changes did not reach statisticalignificance (Table 3). No significant change was notedn the plasma levels of interferon-�, TNF-�, or IL-2, -4,r -5. In contrast, the plasma IL-10 level increased sig-ificantly during G-CSF administration and returned toaseline by 6 weeks (Table 4).

An asymptomatic reversible elevation of alkaline phos-hatase occurred in all patients. Mild bone pain was notedn 2 patients during drug administration. In the 1 patientho did not tolerate the full dose (drug administrationas stopped because of an elevation in alkaline phospha-

ase), renal dysfunction developed secondary to worsen-ng HF; however, the renal function recovered rapidlyith diuretic treatment. None of the patients had implant-

ble defibrillator shocks for ventricular arrhythmias orther cardiovascular adverse events during the 9-monthollow-up period.

• • •obilizing bone marrow-derived HSCs is an attractive con-

ept for acute myocardial infarction and HF. Theoretically,he homing signals that effect HSC engraftment are likely toe greater in acute myocardial infarction, and cell survivalay be optimal in the more vascular failing myocardium

ompared with avascular scar. Engraftment into the myo-ardial scar may cause less electrical heterogeneity thann the HF myocardium. Additional factors to be consid-red in the use of stem cell mobilization are whetheratients with advanced HF can adequately mobilize stemells and whether the neutrophilia induced by hemato-oietic growth factors would be deleterious, especially inhe setting of acute myocardial infarction. Hence, HSCobilization in acute myocardial infarction and HF has

ros and cons; however, the logistics and feasibility, andhe potential public health impact, is greater for patientsith HF.The results of this small pilot study indicate that HSC

an be mobilized in patients with advanced HF withoutignificant adverse effects. Interestingly, the baseline neu-rophil count or hematologic response to G-CSF therapy didot predict the magnitude of HSC mobilization. Although-CSF produced significant neutrophilia, this did not ne-

essitate stopping G-CSF therapy in any patient. A signif-cant effect occurred on left ventricular function in pa-ients with ischemic etiology of advanced HF. Mostnterestingly, G-CSF therapy augmented plasma levels ofhe anti-inflammatory cytokine IL-10, resulting in anncreased anti-inflammatory/pro-inflammatory cytokineatio. Overall, these results suggest the promise of hema-opoietic growth factor therapy in HF and a potentialomplementary mechanism of benefit other than HSCngraftment into the myocardium.

Serum levels of IL-10 are decreased in HF and correlate

egatively with severity,7 and IL-10 significantly inhibits

ridpcGa

niHonpOlncppfi

TB

P

ALEBHSWND

raction

TEh

P

123456M

TEv

P

M

L

L

TEc

C

TIIIII

683Heart Failure/Stem Cell Mobilization and Heart Failure

elease of TNF-� from peripheral blood mononuclear cellssolated from patients with HF.8 G-CSF has been shown toecrease ex vivo TNF-� production in humans throughroduction of IL-10.9 An interesting study by Spuck andoworkers10 examined the effect of a single dose of-CSF, along with antibiotic therapy, compared with

able 1aseline characteristics

atient No. 1 2

ge (yrs) 59 54eft ventricular ejection fraction (%) 24 19tiology Ischemic Ischemiclood pressure (mm Hg) 131/77 122/69eart rate (beats/min) 64 64erum creatinine (mg/dl) 0.9 1.2hite blood cell count (� 103/dl) 6.9 6.22ew York Heart Association class III IIIose (mg/kg) 22.5 25

Mean age � SEM was 63 � 4.2 years; mean left ventricular ejection f

able 2ffects of granulocyte-colony stimulating factor (G-CSF) onematologic parameters

atient No. Baseline Peak

WBC Count(� 103/�l)

CD34� CellCount

(cells/�l)

WBC Count(� 103/�l)

CD34� CellCount

(cells/�l)

6.96 2.9 24.8 20.76.22 5.4 39.9 97.24.16 2.7 73.8 34.8

11.2 5 56.9 49.55.37 2.8 21.6 11.17.82 2.7 35.8 19

ean � SEM 6.9 � 1 3.6 � 0.5 42.1 � 8* 38.7 � 13*

* p �0.05 compared with corresponding baseline group.WBC � white blood cell.

able 3ffects of granulocyte-colony stimulating factor (G-CSF) on leftentricular structure and function

arameter Baseline 9 mo p Value(paired t test)

ean left ventricularejection fraction—allpatients

18.5 � 1.8 24.8 � 3.5 0.069

Ischemic cardiomyopathypatients

20.3 � 2.1 29.5 � 2.9 0.048

eft ventricular end-diastolicdimension—all patients

6.8 � 0.4 6.6 � 0.4 0.517

Ischemic cardiomyopathypatients

6.5 � 0.5 6.1 � 0.4 0.543

eft ventricular end-systolicdimension—all patients

6.3 � 0.3 5.9 � 0.4 0.347

Ischemic cardiomyopathypatients

5.9 � 0.4 5.5 � 0.4 0.474

ntibiotic therapy alone in patients with severe commu-

ity-acquired pneumonia. They demonstrated that G-CSFncreased IL-10 mRNA transcription in neutrophils.ence, it is possible that G-CSF may act by mechanismsther than stem cell transdifferentiation into myocytes, oreovascularization, specifically by altering the cytokinerofile to favor an anti-inflammatory cytokine phenotype.ur results show that plasma pro-inflammatory cytokine

evels were not altered, although G-CSF induced a sig-ificant increase in the levels of the anti-inflammatoryytokine IL-10. It is possible that an alteration in thero-inflammatory/anti-inflammatory cytokine ratio inlasma and possibly myocardium by G-CSF has a bene-cial effect in patients with HF.

1. Orlic D, Kajstura J, Chimenti S, Limana F, Jakoniuk I, Quaini F,Nadal-Ginard B, Bodine DM, Leri A, Anversa P. Mobilized bonemarrow cells repair the infarcted heart, improving function and sur-vival. Proc Natl Acad Sci USA 2001;98:10344–10349.

2. Perin EC, Dohmann HF, Borojevic R, Silva SA, Sousa AL, MesquitaCT, Rossi MI, Carvalho AC, Dutra HS, Dohmann HJ, et al. Transen-docardial, autologous bone marrow cell transplantation for severechronic ischemic heart failure. Circulation 2003;107:2294–2302.

3. Perin EC, Dohmann HF, Borojevic R, Silva SA, Sousa AL, Silva GV,Mesquita CT, Belem L, Vaughn WK, Rangel FO, et al. Improvedexercise capacity and ischemia 6 and 2 months after transendocardialinjection of autologous bone marrow mononuclear cells for ischemiccardiomyopathy. Circulation 2004;110:II213–II218.

4. Ozbaran M, Omay SB, Nalbantgil S, Kultursay H, Kumanlioglu K,Nart D, Pektok E. Autologous peripheral stem cell transplantation inpatients with congestive heart failure due to ischemic heart disease.Eur J Cardiothorac Surg 2004;25:342–350.

5. Weaver CH, Birch R, Greco FA, Schwartzberg L, McAneny B, MooreM, Oviatt D, Redmond J, George C, Alberico T, Johnson P, Buckner

3 4 5 6

60 67 56 8215 23 13 17

Ischemic Ischemic Nonischemic Nonischemic139/98 113/64 91/64 140/8384 82 83 801 1.2 1.1 1.64.16 11.2 5.37 7.82IV III III III

25 22.5 10 25

� SEM was 18.5 � 1.8%.

able 4ffects of granulocyte-colony stimulating factor (G-CSF) on plasmaytokine levels

ytokine Baseline Level (pg/ml) Peak Level (pg/ml) p Value

NF-� 2.8 � 1.5 3.6 � 1.0 NSnterferon-� 14.4 � 13.2 18.8 � 10.3 NSL-2 1.4 � 0.9 1.2 � 0.7 NSL-4 4.8 � 1.0 5.7 � 1.1 NSL-5 2.7 � 0.9 3.1 � 1.0 NSL-10 4.2 � 0.4 9.4 � 1.1 0.003

CD. Mobilization and harvest of peripheral blood stem cells: random-

1

684 The American Journal of Cardiology (www.AJConline.org)

ized evaluations of different doses of filgrastim. Br J Hematol1998;100:338–347.

6. Glaspy JA, Shpall EJ, LeMaistre CF, Briddell RA, Menchaca DM,Turner SA, Lill M, Chap L, Jones R, Wiers MD, Sheridan WP,McNiece IK. Peripheral blood progenitor cell mobilization using stemcell factor in combination with filgrastim in breast cancer patients.Blood 1997;90:2939–2951.

7. Stumpf C, Lehner C, Yilmaz A, Daniel WG, Garlichs CD. Decrease ofserum levels of the anti-inflammatory cytokine interleukin-10 in patientswith advanced chronic heart failure. Clin Sci (Lond) 2003;105:45–50.

8. Bolger AP, Sharma R, von Haehling S, Doehner W, Oliver B, Rauch-

haus M, Coats AJ, Adcock IM, Anker SD. Effect of interleukin-10 onthe production of tumor necrosis factor-alpha by peripheral bloodmononuclear cells from patients with chronic heart failure. Am JCardiol 2002;90:384–389.

9. Agnello D, Mascagni P, Bertini R, Villa P, Senaldi G, Ghezzi P.Granulocyte colony-stimulating factor decreases tumor necrosis factorproduction in whole blood: role of interleukin-10 and prostaglandinE(2). Eur Cytokine Netw 2004;15:323–326.

0. Spuck S, Schaaf B, Wiedorn KH, Hansen F, Vollmer E, Dalhoff K,Braun J. G-CSF application in patients with severe bacterial pneumonia

increases IL-10 expression in neutrophils. Respir Med 2003;97:51–58.